diff options
Diffstat (limited to 'REORG.TODO/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S')
| -rw-r--r-- | REORG.TODO/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S | 806 |
1 files changed, 806 insertions, 0 deletions
diff --git a/REORG.TODO/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S b/REORG.TODO/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S new file mode 100644 index 0000000000..f73ab7de7c --- /dev/null +++ b/REORG.TODO/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S @@ -0,0 +1,806 @@ +/* Function sincosf vectorized with AVX-512. KNL and SKX versions. + Copyright (C) 2014-2017 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + <http://www.gnu.org/licenses/>. */ + +#include <sysdep.h> +#include "svml_s_trig_data.h" +#include "svml_s_wrapper_impl.h" + +/* + ALGORITHM DESCRIPTION: + + 1) Range reduction to [-Pi/4; +Pi/4] interval + a) Grab sign from source argument and save it. + b) Remove sign using AND operation + c) Getting octant Y by 2/Pi multiplication + d) Add "Right Shifter" value + e) Treat obtained value as integer S for destination sign setting. + SS = ((S-S&1)&2)<<30; For sin part + SC = ((S+S&1)&2)<<30; For cos part + f) Change destination sign if source sign is negative + using XOR operation. + g) Subtract "Right Shifter" (0x4B000000) value + h) Subtract Y*(PI/2) from X argument, where PI/2 divided to 4 parts: + X = X - Y*PI1 - Y*PI2 - Y*PI3 - Y*PI4; + 2) Polynomial (minimax for sin within [-Pi/4; +Pi/4] interval) + a) Calculate X^2 = X * X + b) Calculate 2 polynomials for sin and cos: + RS = X * ( A0 + X^2 * (A1 + x^2 * (A2 + x^2 * (A3)))); + RC = B0 + X^2 * (B1 + x^2 * (B2 + x^2 * (B3 + x^2 * (B4)))); + c) Swap RS & RC if if first bit of obtained value after + Right Shifting is set to 1. Using And, Andnot & Or operations. + 3) Destination sign setting + a) Set shifted destination sign using XOR operation: + R1 = XOR( RS, SS ); + R2 = XOR( RC, SC ). */ + + .text +ENTRY (_ZGVeN16vl4l4_sincosf_knl) +#ifndef HAVE_AVX512DQ_ASM_SUPPORT +WRAPPER_IMPL_AVX512_fFF _ZGVdN8vl4l4_sincosf +#else + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $1344, %rsp + movq __svml_s_trig_data@GOTPCREL(%rip), %rax + vmovaps %zmm0, %zmm2 + movl $-1, %edx + vmovups __sAbsMask(%rax), %zmm0 + vmovups __sInvPI(%rax), %zmm3 + +/* Absolute argument computation */ + vpandd %zmm0, %zmm2, %zmm1 + vmovups __sPI1_FMA(%rax), %zmm5 + vmovups __sSignMask(%rax), %zmm9 + vpandnd %zmm2, %zmm0, %zmm0 + +/* h) Subtract Y*(PI/2) from X argument, where PI/2 divided to 3 parts: + X = X - Y*PI1 - Y*PI2 - Y*PI3 */ + vmovaps %zmm1, %zmm6 + vmovaps %zmm1, %zmm8 + +/* c) Getting octant Y by 2/Pi multiplication + d) Add "Right Shifter" value */ + vfmadd213ps __sRShifter(%rax), %zmm1, %zmm3 + vmovups __sPI3_FMA(%rax), %zmm7 + +/* g) Subtract "Right Shifter" (0x4B000000) value */ + vsubps __sRShifter(%rax), %zmm3, %zmm12 + +/* e) Treat obtained value as integer S for destination sign setting */ + vpslld $31, %zmm3, %zmm13 + vmovups __sA7_FMA(%rax), %zmm14 + vfnmadd231ps %zmm12, %zmm5, %zmm6 + +/* 2) Polynomial (minimax for sin within [-Pi/4; +Pi/4] interval) + a) Calculate X^2 = X * X + b) Calculate 2 polynomials for sin and cos: + RS = X * ( A0 + X^2 * (A1 + x^2 * (A2 + x^2 * (A3)))); + RC = B0 + X^2 * (B1 + x^2 * (B2 + x^2 * (B3 + x^2 * (B4)))) */ + vmovaps %zmm14, %zmm15 + vmovups __sA9_FMA(%rax), %zmm3 + vcmpps $22, __sRangeReductionVal(%rax), %zmm1, %k1 + vpbroadcastd %edx, %zmm1{%k1}{z} + vfnmadd231ps __sPI2_FMA(%rax), %zmm12, %zmm6 + vptestmd %zmm1, %zmm1, %k0 + vpandd %zmm6, %zmm9, %zmm11 + kmovw %k0, %ecx + vpxord __sOneHalf(%rax), %zmm11, %zmm4 + +/* Result sign calculations */ + vpternlogd $150, %zmm13, %zmm9, %zmm11 + +/* Add correction term 0.5 for cos() part */ + vaddps %zmm4, %zmm12, %zmm10 + vfnmadd213ps %zmm6, %zmm7, %zmm12 + vfnmadd231ps %zmm10, %zmm5, %zmm8 + vpxord %zmm13, %zmm12, %zmm13 + vmulps %zmm13, %zmm13, %zmm12 + vfnmadd231ps __sPI2_FMA(%rax), %zmm10, %zmm8 + vfmadd231ps __sA9_FMA(%rax), %zmm12, %zmm15 + vfnmadd213ps %zmm8, %zmm7, %zmm10 + vfmadd213ps __sA5_FMA(%rax), %zmm12, %zmm15 + vpxord %zmm11, %zmm10, %zmm5 + vmulps %zmm5, %zmm5, %zmm4 + vfmadd213ps __sA3(%rax), %zmm12, %zmm15 + vfmadd213ps %zmm14, %zmm4, %zmm3 + vmulps %zmm12, %zmm15, %zmm14 + vfmadd213ps __sA5_FMA(%rax), %zmm4, %zmm3 + vfmadd213ps %zmm13, %zmm13, %zmm14 + vfmadd213ps __sA3(%rax), %zmm4, %zmm3 + vpxord %zmm0, %zmm14, %zmm0 + vmulps %zmm4, %zmm3, %zmm3 + vfmadd213ps %zmm5, %zmm5, %zmm3 + testl %ecx, %ecx + jne .LBL_1_3 + +.LBL_1_2: + cfi_remember_state + vmovups %zmm0, (%rdi) + vmovups %zmm3, (%rsi) + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret + +.LBL_1_3: + cfi_restore_state + vmovups %zmm2, 1152(%rsp) + vmovups %zmm0, 1216(%rsp) + vmovups %zmm3, 1280(%rsp) + je .LBL_1_2 + + xorb %dl, %dl + kmovw %k4, 1048(%rsp) + xorl %eax, %eax + kmovw %k5, 1040(%rsp) + kmovw %k6, 1032(%rsp) + kmovw %k7, 1024(%rsp) + vmovups %zmm16, 960(%rsp) + vmovups %zmm17, 896(%rsp) + vmovups %zmm18, 832(%rsp) + vmovups %zmm19, 768(%rsp) + vmovups %zmm20, 704(%rsp) + vmovups %zmm21, 640(%rsp) + vmovups %zmm22, 576(%rsp) + vmovups %zmm23, 512(%rsp) + vmovups %zmm24, 448(%rsp) + vmovups %zmm25, 384(%rsp) + vmovups %zmm26, 320(%rsp) + vmovups %zmm27, 256(%rsp) + vmovups %zmm28, 192(%rsp) + vmovups %zmm29, 128(%rsp) + vmovups %zmm30, 64(%rsp) + vmovups %zmm31, (%rsp) + movq %rsi, 1056(%rsp) + movq %r12, 1096(%rsp) + cfi_offset_rel_rsp (12, 1096) + movb %dl, %r12b + movq %r13, 1088(%rsp) + cfi_offset_rel_rsp (13, 1088) + movl %eax, %r13d + movq %r14, 1080(%rsp) + cfi_offset_rel_rsp (14, 1080) + movl %ecx, %r14d + movq %r15, 1072(%rsp) + cfi_offset_rel_rsp (15, 1072) + movq %rbx, 1064(%rsp) + movq %rdi, %rbx + cfi_remember_state + +.LBL_1_6: + btl %r13d, %r14d + jc .LBL_1_13 + +.LBL_1_7: + lea 1(%r13), %esi + btl %esi, %r14d + jc .LBL_1_10 + +.LBL_1_8: + addb $1, %r12b + addl $2, %r13d + cmpb $16, %r12b + jb .LBL_1_6 + + movq %rbx, %rdi + kmovw 1048(%rsp), %k4 + movq 1056(%rsp), %rsi + kmovw 1040(%rsp), %k5 + movq 1096(%rsp), %r12 + cfi_restore (%r12) + kmovw 1032(%rsp), %k6 + movq 1088(%rsp), %r13 + cfi_restore (%r13) + kmovw 1024(%rsp), %k7 + vmovups 960(%rsp), %zmm16 + vmovups 896(%rsp), %zmm17 + vmovups 832(%rsp), %zmm18 + vmovups 768(%rsp), %zmm19 + vmovups 704(%rsp), %zmm20 + vmovups 640(%rsp), %zmm21 + vmovups 576(%rsp), %zmm22 + vmovups 512(%rsp), %zmm23 + vmovups 448(%rsp), %zmm24 + vmovups 384(%rsp), %zmm25 + vmovups 320(%rsp), %zmm26 + vmovups 256(%rsp), %zmm27 + vmovups 192(%rsp), %zmm28 + vmovups 128(%rsp), %zmm29 + vmovups 64(%rsp), %zmm30 + vmovups (%rsp), %zmm31 + movq 1080(%rsp), %r14 + cfi_restore (%r14) + movq 1072(%rsp), %r15 + cfi_restore (%r15) + movq 1064(%rsp), %rbx + vmovups 1216(%rsp), %zmm0 + vmovups 1280(%rsp), %zmm3 + jmp .LBL_1_2 + +.LBL_1_10: + cfi_restore_state + movzbl %r12b, %r15d + vmovss 1156(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(sinf) + + vmovss %xmm0, 1220(%rsp,%r15,8) + vmovss 1156(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(cosf) + + vmovss %xmm0, 1284(%rsp,%r15,8) + jmp .LBL_1_8 + +.LBL_1_13: + movzbl %r12b, %r15d + vmovss 1152(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(sinf) + + vmovss %xmm0, 1216(%rsp,%r15,8) + vmovss 1152(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(cosf) + + vmovss %xmm0, 1280(%rsp,%r15,8) + jmp .LBL_1_7 +#endif +END (_ZGVeN16vl4l4_sincosf_knl) +libmvec_hidden_def(_ZGVeN16vl4l4_sincosf_knl) + +ENTRY (_ZGVeN16vl4l4_sincosf_skx) +#ifndef HAVE_AVX512DQ_ASM_SUPPORT +WRAPPER_IMPL_AVX512_fFF _ZGVdN8vvv_sincosf +#else + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $1344, %rsp + movq __svml_s_trig_data@GOTPCREL(%rip), %rax + vmovaps %zmm0, %zmm4 + vmovups __sAbsMask(%rax), %zmm3 + vmovups __sInvPI(%rax), %zmm5 + vmovups __sRShifter(%rax), %zmm6 + vmovups __sPI1_FMA(%rax), %zmm9 + vmovups __sPI2_FMA(%rax), %zmm10 + vmovups __sSignMask(%rax), %zmm14 + vmovups __sOneHalf(%rax), %zmm7 + vmovups __sPI3_FMA(%rax), %zmm12 + +/* Absolute argument computation */ + vandps %zmm3, %zmm4, %zmm2 + +/* c) Getting octant Y by 2/Pi multiplication + d) Add "Right Shifter" value */ + vfmadd213ps %zmm6, %zmm2, %zmm5 + vcmpps $18, __sRangeReductionVal(%rax), %zmm2, %k1 + +/* e) Treat obtained value as integer S for destination sign setting */ + vpslld $31, %zmm5, %zmm0 + +/* g) Subtract "Right Shifter" (0x4B000000) value */ + vsubps %zmm6, %zmm5, %zmm5 + vmovups __sA3(%rax), %zmm6 + +/* h) Subtract Y*(PI/2) from X argument, where PI/2 divided to 3 parts: + X = X - Y*PI1 - Y*PI2 - Y*PI3 */ + vmovaps %zmm2, %zmm11 + vfnmadd231ps %zmm5, %zmm9, %zmm11 + vfnmadd231ps %zmm5, %zmm10, %zmm11 + vandps %zmm11, %zmm14, %zmm1 + vxorps %zmm1, %zmm7, %zmm8 + +/* Result sign calculations */ + vpternlogd $150, %zmm0, %zmm14, %zmm1 + vmovups .L_2il0floatpacket.13(%rip), %zmm14 + +/* Add correction term 0.5 for cos() part */ + vaddps %zmm8, %zmm5, %zmm15 + vfnmadd213ps %zmm11, %zmm12, %zmm5 + vandnps %zmm4, %zmm3, %zmm11 + vmovups __sA7_FMA(%rax), %zmm3 + vmovaps %zmm2, %zmm13 + vfnmadd231ps %zmm15, %zmm9, %zmm13 + vxorps %zmm0, %zmm5, %zmm9 + vmovups __sA5_FMA(%rax), %zmm0 + vfnmadd231ps %zmm15, %zmm10, %zmm13 + vmulps %zmm9, %zmm9, %zmm8 + vfnmadd213ps %zmm13, %zmm12, %zmm15 + vmovups __sA9_FMA(%rax), %zmm12 + vxorps %zmm1, %zmm15, %zmm1 + vmulps %zmm1, %zmm1, %zmm13 + +/* 2) Polynomial (minimax for sin within [-Pi/4; +Pi/4] interval) + a) Calculate X^2 = X * X + b) Calculate 2 polynomials for sin and cos: + RS = X * ( A0 + X^2 * (A1 + x^2 * (A2 + x^2 * (A3)))); + RC = B0 + X^2 * (B1 + x^2 * (B2 + x^2 * (B3 + x^2 * (B4)))) */ + vmovaps %zmm12, %zmm7 + vfmadd213ps %zmm3, %zmm8, %zmm7 + vfmadd213ps %zmm3, %zmm13, %zmm12 + vfmadd213ps %zmm0, %zmm8, %zmm7 + vfmadd213ps %zmm0, %zmm13, %zmm12 + vfmadd213ps %zmm6, %zmm8, %zmm7 + vfmadd213ps %zmm6, %zmm13, %zmm12 + vmulps %zmm8, %zmm7, %zmm10 + vmulps %zmm13, %zmm12, %zmm3 + vfmadd213ps %zmm9, %zmm9, %zmm10 + vfmadd213ps %zmm1, %zmm1, %zmm3 + vxorps %zmm11, %zmm10, %zmm0 + vpandnd %zmm2, %zmm2, %zmm14{%k1} + vptestmd %zmm14, %zmm14, %k0 + kmovw %k0, %ecx + testl %ecx, %ecx + jne .LBL_2_3 + +.LBL_2_2: + cfi_remember_state + vmovups %zmm0, (%rdi) + vmovups %zmm3, (%rsi) + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret + +.LBL_2_3: + cfi_restore_state + vmovups %zmm4, 1152(%rsp) + vmovups %zmm0, 1216(%rsp) + vmovups %zmm3, 1280(%rsp) + je .LBL_2_2 + + xorb %dl, %dl + xorl %eax, %eax + kmovw %k4, 1048(%rsp) + kmovw %k5, 1040(%rsp) + kmovw %k6, 1032(%rsp) + kmovw %k7, 1024(%rsp) + vmovups %zmm16, 960(%rsp) + vmovups %zmm17, 896(%rsp) + vmovups %zmm18, 832(%rsp) + vmovups %zmm19, 768(%rsp) + vmovups %zmm20, 704(%rsp) + vmovups %zmm21, 640(%rsp) + vmovups %zmm22, 576(%rsp) + vmovups %zmm23, 512(%rsp) + vmovups %zmm24, 448(%rsp) + vmovups %zmm25, 384(%rsp) + vmovups %zmm26, 320(%rsp) + vmovups %zmm27, 256(%rsp) + vmovups %zmm28, 192(%rsp) + vmovups %zmm29, 128(%rsp) + vmovups %zmm30, 64(%rsp) + vmovups %zmm31, (%rsp) + movq %rsi, 1056(%rsp) + movq %r12, 1096(%rsp) + cfi_offset_rel_rsp (12, 1096) + movb %dl, %r12b + movq %r13, 1088(%rsp) + cfi_offset_rel_rsp (13, 1088) + movl %eax, %r13d + movq %r14, 1080(%rsp) + cfi_offset_rel_rsp (14, 1080) + movl %ecx, %r14d + movq %r15, 1072(%rsp) + cfi_offset_rel_rsp (15, 1072) + movq %rbx, 1064(%rsp) + movq %rdi, %rbx + cfi_remember_state + +.LBL_2_6: + btl %r13d, %r14d + jc .LBL_2_13 + +.LBL_2_7: + lea 1(%r13), %esi + btl %esi, %r14d + jc .LBL_2_10 + +.LBL_2_8: + incb %r12b + addl $2, %r13d + cmpb $16, %r12b + jb .LBL_2_6 + + kmovw 1048(%rsp), %k4 + movq %rbx, %rdi + kmovw 1040(%rsp), %k5 + kmovw 1032(%rsp), %k6 + kmovw 1024(%rsp), %k7 + vmovups 960(%rsp), %zmm16 + vmovups 896(%rsp), %zmm17 + vmovups 832(%rsp), %zmm18 + vmovups 768(%rsp), %zmm19 + vmovups 704(%rsp), %zmm20 + vmovups 640(%rsp), %zmm21 + vmovups 576(%rsp), %zmm22 + vmovups 512(%rsp), %zmm23 + vmovups 448(%rsp), %zmm24 + vmovups 384(%rsp), %zmm25 + vmovups 320(%rsp), %zmm26 + vmovups 256(%rsp), %zmm27 + vmovups 192(%rsp), %zmm28 + vmovups 128(%rsp), %zmm29 + vmovups 64(%rsp), %zmm30 + vmovups (%rsp), %zmm31 + vmovups 1216(%rsp), %zmm0 + vmovups 1280(%rsp), %zmm3 + movq 1056(%rsp), %rsi + movq 1096(%rsp), %r12 + cfi_restore (%r12) + movq 1088(%rsp), %r13 + cfi_restore (%r13) + movq 1080(%rsp), %r14 + cfi_restore (%r14) + movq 1072(%rsp), %r15 + cfi_restore (%r15) + movq 1064(%rsp), %rbx + jmp .LBL_2_2 + +.LBL_2_10: + cfi_restore_state + movzbl %r12b, %r15d + vmovss 1156(%rsp,%r15,8), %xmm0 + vzeroupper + vmovss 1156(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(sinf) + + vmovss %xmm0, 1220(%rsp,%r15,8) + vmovss 1156(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(cosf) + + vmovss %xmm0, 1284(%rsp,%r15,8) + jmp .LBL_2_8 + +.LBL_2_13: + movzbl %r12b, %r15d + vmovss 1152(%rsp,%r15,8), %xmm0 + vzeroupper + vmovss 1152(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(sinf) + + vmovss %xmm0, 1216(%rsp,%r15,8) + vmovss 1152(%rsp,%r15,8), %xmm0 + + call JUMPTARGET(cosf) + + vmovss %xmm0, 1280(%rsp,%r15,8) + jmp .LBL_2_7 +#endif +END (_ZGVeN16vl4l4_sincosf_skx) +libmvec_hidden_def(_ZGVeN16vl4l4_sincosf_skx) + +/* Wrapper between vvv and vl4l4 vector variants. */ +.macro WRAPPER_AVX512_vvv_vl4l4 callee +#ifndef __ILP32__ + pushq %rbp + cfi_adjust_cfa_offset (8) + cfi_rel_offset (%rbp, 0) + movq %rsp, %rbp + cfi_def_cfa_register (%rbp) + andq $-64, %rsp + subq $384, %rsp + /* Encoding for vmovups %zmm1, 128(%rsp). */ + .byte 0x62 + .byte 0xf1 + .byte 0x7c + .byte 0x48 + .byte 0x11 + .byte 0x4c + .byte 0x24 + .byte 0x02 + lea (%rsp), %rdi + /* Encoding for vmovups %zmm2, 192(%rdi). */ + .byte 0x62 + .byte 0xf1 + .byte 0x7c + .byte 0x48 + .byte 0x11 + .byte 0x57 + .byte 0x03 + /* Encoding for vmovups %zmm3, 256(%rdi). */ + .byte 0x62 + .byte 0xf1 + .byte 0x7c + .byte 0x48 + .byte 0x11 + .byte 0x5f + .byte 0x04 + /* Encoding for vmovups %zmm4, 320(%rdi). */ + .byte 0x62 + .byte 0xf1 + .byte 0x7c + .byte 0x48 + .byte 0x11 + .byte 0x67 + .byte 0x05 + lea 64(%rsp), %rsi + call HIDDEN_JUMPTARGET(\callee) + movq 128(%rsp), %rdx + movq 136(%rsp), %rsi + movq 144(%rsp), %r8 + movq 152(%rsp), %r10 + movl (%rsp), %eax + movl 4(%rsp), %ecx + movl 8(%rsp), %edi + movl 12(%rsp), %r9d + movl %eax, (%rdx) + movl %ecx, (%rsi) + movq 160(%rsp), %rax + movq 168(%rsp), %rcx + movl %edi, (%r8) + movl %r9d, (%r10) + movq 176(%rsp), %rdi + movq 184(%rsp), %r9 + movl 16(%rsp), %r11d + movl 20(%rsp), %edx + movl 24(%rsp), %esi + movl 28(%rsp), %r8d + movl %r11d, (%rax) + movl %edx, (%rcx) + movq 192(%rsp), %r11 + movq 200(%rsp), %rdx + movl %esi, (%rdi) + movl %r8d, (%r9) + movq 208(%rsp), %rsi + movq 216(%rsp), %r8 + movl 32(%rsp), %r10d + movl 36(%rsp), %eax + movl 40(%rsp), %ecx + movl 44(%rsp), %edi + movl %r10d, (%r11) + movl %eax, (%rdx) + movq 224(%rsp), %r10 + movq 232(%rsp), %rax + movl %ecx, (%rsi) + movl %edi, (%r8) + movq 240(%rsp), %rcx + movq 248(%rsp), %rdi + movl 48(%rsp), %r9d + movl 52(%rsp), %r11d + movl 56(%rsp), %edx + movl 60(%rsp), %esi + movl %r9d, (%r10) + movl %r11d, (%rax) + movq 256(%rsp), %r9 + movq 264(%rsp), %r11 + movl %edx, (%rcx) + movl %esi, (%rdi) + movq 272(%rsp), %rdx + movq 280(%rsp), %rsi + movl 64(%rsp), %r8d + movl 68(%rsp), %r10d + movl 72(%rsp), %eax + movl 76(%rsp), %ecx + movl %r8d, (%r9) + movl %r10d, (%r11) + movq 288(%rsp), %r8 + movq 296(%rsp), %r10 + movl %eax, (%rdx) + movl %ecx, (%rsi) + movq 304(%rsp), %rax + movq 312(%rsp), %rcx + movl 80(%rsp), %edi + movl 84(%rsp), %r9d + movl 88(%rsp), %r11d + movl 92(%rsp), %edx + movl %edi, (%r8) + movl %r9d, (%r10) + movq 320(%rsp), %rdi + movq 328(%rsp), %r9 + movl %r11d, (%rax) + movl %edx, (%rcx) + movq 336(%rsp), %r11 + movq 344(%rsp), %rdx + movl 96(%rsp), %esi + movl 100(%rsp), %r8d + movl 104(%rsp), %r10d + movl 108(%rsp), %eax + movl %esi, (%rdi) + movl %r8d, (%r9) + movq 352(%rsp), %rsi + movq 360(%rsp), %r8 + movl %r10d, (%r11) + movl %eax, (%rdx) + movq 368(%rsp), %r10 + movq 376(%rsp), %rax + movl 112(%rsp), %ecx + movl 116(%rsp), %edi + movl 120(%rsp), %r9d + movl 124(%rsp), %r11d + movl %ecx, (%rsi) + movl %edi, (%r8) + movl %r9d, (%r10) + movl %r11d, (%rax) + movq %rbp, %rsp + cfi_def_cfa_register (%rsp) + popq %rbp + cfi_adjust_cfa_offset (-8) + cfi_restore (%rbp) + ret +#else + leal 8(%rsp), %r10d + .cfi_def_cfa 10, 0 + andl $-64, %esp + pushq -8(%r10d) + pushq %rbp + .cfi_escape 0x10,0x6,0x2,0x76,0 + movl %esp, %ebp + pushq %r10 + .cfi_escape 0xf,0x3,0x76,0x78,0x6 + leal -112(%rbp), %esi + leal -176(%rbp), %edi + subl $296, %esp + /* Encoding for vmovdqa64 %zmm1, -240(%ebp). */ + .byte 0x67 + .byte 0x62 + .byte 0xf1 + .byte 0xfd + .byte 0x48 + .byte 0x7f + .byte 0x8d + .byte 0x10 + .byte 0xff + .byte 0xff + .byte 0xff + /* Encoding for vmovdqa64 %zmm2, -304(%ebp). */ + .byte 0x67 + .byte 0x62 + .byte 0xf1 + .byte 0xfd + .byte 0x48 + .byte 0x7f + .byte 0x95 + .byte 0xd0 + .byte 0xfe + .byte 0xff + .byte 0xff + call HIDDEN_JUMPTARGET(\callee) + movl -240(%ebp), %eax + vmovss -176(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -236(%ebp), %eax + vmovss -172(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -232(%ebp), %eax + vmovss -168(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -228(%ebp), %eax + vmovss -164(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -224(%ebp), %eax + vmovss -160(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -220(%ebp), %eax + vmovss -156(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -216(%ebp), %eax + vmovss -152(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -212(%ebp), %eax + vmovss -148(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -208(%ebp), %eax + vmovss -144(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -204(%ebp), %eax + vmovss -140(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -200(%ebp), %eax + vmovss -136(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -196(%ebp), %eax + vmovss -132(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -192(%ebp), %eax + vmovss -128(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -188(%ebp), %eax + vmovss -124(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -184(%ebp), %eax + vmovss -120(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -180(%ebp), %eax + vmovss -116(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -304(%ebp), %eax + vmovss -112(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -300(%ebp), %eax + vmovss -108(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -296(%ebp), %eax + vmovss -104(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -292(%ebp), %eax + vmovss -100(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -288(%ebp), %eax + vmovss -96(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -284(%ebp), %eax + vmovss -92(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -280(%ebp), %eax + vmovss -88(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -276(%ebp), %eax + vmovss -84(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -272(%ebp), %eax + vmovss -80(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -268(%ebp), %eax + vmovss -76(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -264(%ebp), %eax + vmovss -72(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -260(%ebp), %eax + vmovss -68(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -256(%ebp), %eax + vmovss -64(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -252(%ebp), %eax + vmovss -60(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -248(%ebp), %eax + vmovss -56(%ebp), %xmm0 + vmovss %xmm0, (%eax) + movl -244(%ebp), %eax + vmovss -52(%ebp), %xmm0 + vmovss %xmm0, (%eax) + addl $296, %esp + popq %r10 + .cfi_def_cfa 10, 0 + popq %rbp + leal -8(%r10), %esp + .cfi_def_cfa 7, 8 + ret +#endif +.endm + +ENTRY (_ZGVeN16vvv_sincosf_knl) +WRAPPER_AVX512_vvv_vl4l4 _ZGVeN16vl4l4_sincosf_knl +END (_ZGVeN16vvv_sincosf_knl) + +ENTRY (_ZGVeN16vvv_sincosf_skx) +WRAPPER_AVX512_vvv_vl4l4 _ZGVeN16vl4l4_sincosf_skx +END (_ZGVeN16vvv_sincosf_skx) + + .section .rodata, "a" +.L_2il0floatpacket.13: + .long 0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff,0xffffffff + .type .L_2il0floatpacket.13,@object |